In many data processing applications records are made of events, and these records are then stored in a database. A user may then query the database to obtain a visual display of the information relating to the stored records.
One type of database application that is of particular interest to this invention is one that measures and records data processing system performance. In some of these systems the data for a number of monitored computers is maintained in a central database. The data typically includes performance metrics, such as CPU Utilization, Disk Utilization, Memory Utilization, Input/Output (I/O) Wait Time, and other similar performance-related metrics.
It is often the case that data is stored with a higher resolution for more recent periods of time. For example, the metrics can be displayed for one day, but only for the past 30 days; and the metrics can be displayed for one week, but only for the past 52 weeks; and the metrics can be displayed for one month, but only for the past three years. This arrangement avoids a requirement to save data at a high resolution (e.g., once per day) for longer than is necessary, and thus conserves system memory. The underlying assumption in this type of data storage scheme is that long-term analysis will only be concerned with major trends, whereas short-term analysis will only be concerned with recent specific events and conditions. In this scheme the older data can be said to be “rolled-up”, that is, averaged over longer periods of time (e.g., performance metrics for each day of a week are averaged into one performance metric for the entire week).
During use, a user will typically review a monthly report, and then attempt to view a particular week within the month. However, this will be effective only if the month of interest occurred within the past year. Likewise, a user viewing a weekly report may wish to view a particular day within a week of interest, but may find that the week of interest did not occur recently enough for the individual days to be available for viewing.
As such, a need exists for a user interface that enables a user to quickly and efficiently determine, at a glance, what time periods, such as months, weeks and days, are currently available to be viewed, and what relationships exist between the available time periods.
In addition, and in order to avoid the user having to generate and review many reports, a need exists to provide a user interface that enables a user to determine, at a glance, what months, weeks and days have data that indicate a problem, such as a system bottleneck or a system outage.
Representative U.S. patents that may be interest to this invention include the following.
U.S. Pat. No. 6,397,359 B1, “Methods, Systems and Computer Program Products for Scheduled Network Performance Testing”, V. Chandra et al., describes a system to test network performance by defining test schedules to be implemented, and when test protocols should be executed. In this system it is said to not be necessary to continuously update test-related auto-thresholds, particularly if sufficient damping is provided, to insure stable operations of performance analysis and threshold crossing detection. Auto-thresholds may be adjusted for those connections which have auto-thresholds on a periodic basis, such as a nightly basis. This time period may be increased as greater amounts of prior data are obtained and are available for a given connection and threshold. As an example, for a connection and threshold having a full week or more of data, the auto-threshold may only be updated on a weekly basis, rather than on a nightly basis
Commonly assigned U.S. Pat. No. 6,470,464, “System and Method for Predicting Computer System Performance and for Making Recommendations for Improving Its Performance”, R. L. Bertram et al., describes a system and method of monitoring and analyzing the performance of a computer system and its components in a data processing network, and for proposing changes to the network to improve the performance. The system involves identifying undesirable conditions (sometimes called bottlenecks) in the system, determining which bottlenecks are the most severe in affecting the performance of the system, and proposing changes to the components of the system to improve performance of the network. This U.S. patent uses historical data and forecasting techniques to predict bottlenecks which have not yet occurred, but which can be expected, so that further bottlenecks can be projected, along with a prediction interval to indicate the confidence of the prediction. The further projections and the prediction interval are provided in a graph that is available to the user.
In addition, U.S. Pat. No. 5,093,794, “Job Scheduling System”, G. R. Howie et al., describes a system to schedule jobs by the use of a time map that also indicates bottlenecks. JP09113308, “Measurement Data Processor”, R. Hagiwara et al., describes, in the Abstract, a raw data file into which raw data, obtained by averaging measurement data, are written at one minute intervals; a daily report data file into which the sum value or average value of the raw data, per one hour, are written; a monthly report data file into which the sum value of the average value of daily report data, per one day, are written; and an annual report data file into which the sum value or average value of the monthly report data, per one year, are written. In response to a command to display a graph, a processor reads out the necessary data from a required file and displays the graph.
As should be apparent, prior to this invention the foregoing and other user interface needs have not been adequately fulfilled.